Female marriage age, education, income, and health B1.1 Existing evidence: early marriage and female and child health In much of the developing worid, early female marriagedefined as marriage before the age of 18 remains widespread despite age of consent laws banning the practice, government and NGO efforts to curtail it, increasing educafion levels, and economic growth (National Research Council and Institute of Medicine, 2005). A recent study by UNICEF revealed that in Latin America and the Caribbean 29 percent of women were married by the fime they turned 18; in Africa, 42 percent; and in Southern Asia, 48 percent (UNICEF, 2005). Bangladesh has one of the highest rates of adolescent and child marriage in the worid: Although the legal age of marriage for females is 18, nearly 50 percent of all girls and 75 percent of rural giris are estimated to be married by age 15 (UNICEF 2006). There is substantial literature showing a correlation between eariy marriage and women's health, [and health-seeking behavior]. In general, women who marry early begin childbearing at a young age (Jensen and Thornton, 2003), and complicafions in pregnancy and delivery are a leading cause of death among giris aged 15 to 19. Maternal mortality in this group is double the rates for women in their 20s. Giris who marry as adolescents face greater health risks associated with lower age of first birth, higher fertility rates, and shorter birth spacing related to lower contraceptive use (UNICEF, 2001). About 60 percent of adolescent giris in Bangladesh are mothers by age 19, and nearly all of them married (UNICEF, 2006). The 2004 Demographic Health Survey (DHS) found that 40.7 percent become mothers between 15 and 17 years, and 19.5 percent between ages 18 and 19 (NIPORT, Mitra Associates and ORC Macro, 2005). Childbearing during adolescence, when physiology is likely to be underdeveloped, is widely believed to result in higher levels of maternal mortality and morbidity, although the degree to which age infiuences reproductive outcomes is not well established. Giris age 14 and younger are five times as likely to die from pregnancy complications and their offspring are also less likely to survive (UNFPA, 2004). The discrepancy in maternal mortality by age persists even in countries where maternal mortality is low, such as Brazil and the United States (Mathur, et al., 2003). In Bangladesh, maternal mortality and severe morbidity from childbirth is neariy twice as high and rates of postnatal care are 50 percent lower for adolescent giris compared to women ages 20 to 35. Furthermore, married women ages 15 to 19 are less likely to use modern contraceptives than married women ages 20 to 24 (Population Reference Bureau. 2006). Young mothers also have higher maternal morbidity rates, including severe complicafions, such as obstructed labor or obstetric fistula, which occur primarily among young women (UNFPA and EngenderHealth, 2003; Jarrett, 1994). Without fimely intervention obstructed labor can lead to tissue necrosis, which can result in permanent maternal morbidity, if not mortality. Data on maternal morbidity is scarce, and only available for a handful of settings. In Ethiopia, where 24 percent of women give birth by age 18, obstructed labor is the immediate cause in 46 percent of maternal deaths, and three in 1,000 pregnant women develop fistula, which is also common in Bangladesh (Populafion Reference Bureau, 2006; UNFPA, 2003; Akhter, et al., 1996). Of fundamental importance is the fact that the medical community currenfiy does not know the degree to which the well-documented relafionships between age of childbearing and reproducfive outcomes are physiological consequences of eariy childbearing. Hence, this research has important scienfific value. In addition to the physiological channels, eariy marriage may also impact health through behavioral channels. First, youth is associated with less-active health-seeking behavior and limited health information, which has a negative impact on the health status of married adolescent giris. In Bangladesh, 70 percent of pregnant giris younger than 20 receive no antenatal care and 90 percent deliver their babies at home. Their access to health information is poor: 20 percent of adolescent mothers have little knowledge of life-threatening conditions during pregnancy, and the majority (married and unmarried) have no informafion on sexuality, contraception, or sexually transmitted infecfions or HIV/AIDS (Haider, et al., 1997; Nahar, et al., 1999; Barkat, et al., 2000; Bruce and Clark, 2004). Adolescent giris' access to reproductive health care and services is also poor: In Bangladesh, the need for contraception is not met for 27 percent of mothers below age 20, compared with 10 percent among those aged 20 to 35 (NIPORT, Mitra Associates and ORC Macro, 2001). Moreover, married adolescents use contraception at much lower rates than older women. In South Asia, 9 percent of married women ages 15 to 19 use modern contraception compared to 24 percent of women ages 20 to 24. In Bangladesh, the rates are 34 percent and 47 percent, respectively (Population Reference Bureau, 2006). Lower usage may refiect lack of awareness about family planning, expectations to have the first child immediately, and more limited access to health services among adolescents. In addition, younger girls tend to marry significantly older men. Research in sub-Saharan Africa found that the husbands of giris ages 15 to 19 years are on average 10 years older (UNICEF 2001). Mean spouse age difference is decreasing with women's age at first marriage throughout the worid. In West Africa, the mean spouse age difference is 12 years for girls aged 14 to 15 at first marriage, and 8 years for women married at 24 to 25 years; the same pattern is found in Southern Asia (UNFPA 2004). The presence of a large age gap between spouses can contribute to poor outcomes in a number of ways. First, older husbands tend to be more sexually experienced, which implies greater risk of sexually transmitted infecfion (Clark, 2004; Luke and Kurz, 2002). The age gap is also associated with lack of agency in marriage for the adolescent giri, which may contribute to poor health outcomes. Lack of decision-making power may translate into lower reproducfive control, or capacity to negofiate sexual relations, contracepfion, and childbearing. Qualitative research also suggests that most young married giris face pressure to get pregnant eariy in marriage and lack reproductive control to avoid it (Bledsoe and Cohen, 1993; Mensch, Bruce, and Greene, 1998; Bruce and Clark, 2004). There is qualitative but little rigorous analysis suggesfing that isolation, restricted mobility, and lack of control over household resources are more common among young married giris (Mensch, et al., 1998). Isolation and the increased stress of adult responsibilifies may have a direct detrimental impact on psychological health. Lack of mobility is also likely to contribute to low healthcare utilization among married adolescent giris. Research in India has documented that married adolescent giris' healthcare decisions are mostly controlled by husbands and mothers-in-law (Barua and Kurz, 2001). Taken together with restricted mobility, this may limit the ability of adolescent giris to access health services for themselves and their children. Finally, the negafive associafion between eariy marriage and health extends to the next generation. Children born to women under age 20 have higher infant mortality rates (IMR) through the age of five. In Mali, the Infant mortality rate is 181 per 1,000 for children of mothers under 20 compared to 111 per 1,000 for children of mothers aged 20 to 29. Similariy, these rates are 164 and 88 in Tanzania, 108 and 68 in Nepal, and 71 and 28 in the Dominican Republic (Marthur, Green, and Malhotra, 2003). In Bangladesh, the IMR is 86 for infants born to mother under 20 compared to 60 for mothers aged 20 to 29 (NIPORT, Mitra Associates and ORC Macro, 2005). Child mortality rates (CMR) are also higher for children of adolescent mothers. In Kenya, the rate is 48 per 1,000 for children born to mothers under 20 compared to 32 for children born to mothers aged 20 to 29. Comparable figures are 90 and 83 in Ethiopia; 40 and 19 in South Africa; and 15 and 13 in Egypt (Marthur, Green, and Malhotra, 2003). In Bangladesh, the CMR is 106 per 1,000 for children of mothers under 20 compared to 84 for children born to mothers aged 20 to 29 (NIPORT, Mitra Associates and ORC Macro, 2005). [How much of this correlation is due to lower utilization of health care (e.g., lower immunization rates) or less knowledge of good health practices by mothers on the part of children is unclear.]